University of Warwick institutional repository:

Transcription

1 University of Warwick institutional repository: This paper is made available online in accordance with publisher policies. Please scroll down to view the document itself. Please refer to the repository record for this item and our policy information available from the repository home page for further information. To see the final version of this paper please visit the publisher s website. Access to the published version may require a subscription. Author(s): Z. Tamainot-Telto, S.J. Metcalf, R.E. Critoph, Y. Zhong, R. Thorpe Article Title: Carbon ammonia pairs for adsorption refrigeration applications: ice making, air conditioning and heat pumping Year of publication: 9 Link to published article: Publisher statement: NOTICE: this is the author s version of a work that was accepted for publication in International Journal of Refrigeration. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in International Journal of Refrigeration, [VOL:3, ISSUE:6, September 9, DOI:.6/j.ijrefrig.9..8,

2 Carbon-Ammonia pairs for adsorption refrigeration applications: ice making, air conditioning and heat pumping Z. Tamainot-Telto (), S.J. Metcalf, R.E. Critoph, Y. Zhong., R. Thorpe School of Engineering - University of Warwick Coventry CV4 7AL United Kingdom (UK) Abstract A thermodynamic cycle model is used to select an optimum adsorbent-refrigerant pair in respect of a chosen figure of merit that could be the cooling production (MJ m -3 ), the heating production (MJ m -3 ) or the coefficient of performance (). This model is based mainly on the adsorption equilibrium equations of the adsorbent-refrigerant pair and heat flows. The simulation results of 6 various activated carbon-ammonia pairs for three cycles (single bed, two-bed and infinite number of beds) are presented at typical conditions for ice making, air conditioning and heat pumping applications. The driving temperature varies from 8 C to C.. The carbon absorbents investigated are mainly coconut shell and coal based types in multiple forms: monolithic, granular, compacted granular, fibre, compacted fibre, cloth, compacted cloth and powder. Considering a two-bed cycle, the best thermal performances based on power density are obtained with the monolithic carbon KOH-AC.with a driving temperature of C, the cooling production is about 66 MJ m -3 (=.45) and 5 MJ m -3 (=.6) for ice making and air conditioning respectively; the heating production is about 36 MJ m -3 (=.5). (Keywords: Adsorption, Air conditioning, Ammonia, Carbon, Heat pump, Ice making, Refrigeration) Les paires charbon actif-ammoniac pour les applications de réfrigération à adsorption: fabrication de la glace, climatisation et pompe à chaleur Un modèle du cycle thermodynamique est utilisé pour sélectionner le meilleur couple adsorbant-ammoniac sur la base de la production frigorifique (MJ m -3 ), la production de chaleur (MJ m -3 ) ou bien le coefficient de performance (). Ce modèle est essentiellement basé sur les équations d état de l adsorption (adsorbant-ammoniac). Les résultats de simulation de 6 différentes paires charbon actif-ammoniac sont présentés pour des conditions typiques de fabrication de la glace, de climatisation et de pompe à chaleur. La température de génération varie de 8 C à C. Les simulations sont effectuées pour trois types de cycle: lit unique, deux-lits et un nombre infini de lits. Les charbons actifs étudiés sont essentiellement issus des coquilles de noix de coco et de minerais de charbon sous diverses formes: monolithique, granulaire, granulaire compacté, fibre, fibre compactée, tissu, tissu compacté et poudre. En considérant un cycle à deux lits, les meilleures performances thermiques sont obtenues avec le charbon monolithique KOH-AC:avec une température de génération de C: la production frigorifique est environ 66 MJ m -3 (=.45) et 5 MJ m -3 (=.6) pour la fabrication de la glace et la climatisation respectivement; la production de chaleur est environ 36 MJ m -3 (=.5). (Mots clés: Adsorption, Conditionnement d air, Ammoniac, Charbon, Pompe à chaleur, Machine à glace, Réfrigération) () Corresponding author: es7@eng.warwick.ac.uk - Tel Fax

3 NOMENCLATURE A Ammonia characteristic coefficient (K) C Specific heat (J kg - K - ) Coefficient of performance h Specific enthalpy (J kg - ) H Specific heat of sorption (J kg - ) k n Dubinin coefficient Dubinin coefficient q Heat density (J m -3 ) Q Heat (J kg - ) P Pressure (Pas or Bar) R Ammonia gas constant (J kg - K - ) SEE Standard Estimated Error T Temperature (K or o C) TR x Tons of refrigeration Concentration (kg Ammonia kg - Carbon) Greek letters ρ Density (kg m- 3 ) Difference, variation Subscripts a c C E G H In p reg reject adsorbed phase carbon, cooling Condensing, condensation Evaporating, evaporation Generating, generator Heating Input At constant pressure Regeneration Rejection, Rejecting

4 sat tip Under ammonia saturation conditions Tipping 3

5 INTRODUCTION & BACKGROUND The basic adsorption refrigeration system consists of two linked vessels, one of which contains the adsorbent-refrigerant pair (generator) and the second which contains only refrigerant (evaporator-condenser). Both vessels are initially at low pressure and temperature with a high refrigerant concentration within the adsorbent and only refrigerant gas in the second vessel (Figure ). The first step consists of heating up the generator using the chosen or available heat source: the refrigerant gas is driven out from the adsorbent while the pressure of the full system rises (desorption) (a). The desorbed gas is condensed in the evaporator-condenser by rejecting heat (b). When the generator has reached the minimum or desirable refrigerant concentration (c), it is then cooled to its initial temperature and readsorbs the refrigerant, reducing the pressure (adsorption). The lower pressure causes the refrigerant liquid contained in the evaporator-condenser to boil, absorbing heat and therefore producing the necessary cooling effect (d). The basic adsorption refrigeration cycle is intermittent and the cooling production takes place during the half cycle only. However, two or more beds can be operated out of phase to produce continuous cooling. The basic thermodynamic cycle of an adsorption machine is described in Figure. An adsorption cycle is characterised by four operating temperatures: the generating temperature (T 3 ), the initial or lower generator temperature (T ), the condensing temperature (T C ) and the evaporating temperature (T E ). (d) P P (b) P P P T (a) T T (c) T T ( a ) Ammonia Ammonia a Gas Gas Cool adsorbent Condensation Hot heat rejected adsorbent Heat Adsorption Refrigeration heat rejected (heat in) Cool adsorbent Figure : Principle of adsorption refrigeration technology T T 3 T C 3 T T 4 T E 4 Figure : Adsorption thermodynamic cycle 4

6 to and to 3: Isosteric and isobaric heat input respectively (Desorption process)3 to 4 and 4 to : Isosteric and isobaric cooling (rejected heat) respectively (Adsorption process) Unlike a conventional refrigeration system driven mechanically by a compressor, the adsorption refrigeration system is a heat driven machine. Since that heat could be widely available or from a waste source, adsorption refrigeration systems offer a great advantage compared with conventional systems. The use of such systems (powered by solar energy or heat generated by burning agricultural waste or biomass in general) in the remote part of developing countries where there is no electricity supply is useful for storage of medical products (vaccine), foods (vegetable, meat, fish, etc.) or habitat comfort (air conditioning). For the food conservation application, the basic adsorption system could provide up to. TR (equivalent of kg ice/day) with low generating temperature ( C < T 3 < C) []. The basic solar sorption technology is not yet as cost effective as the conventional vapour compression cooling system because of its poor specific cooling production: with activated carbon, the typical specific cooling production is less than W kg - carbon [-7] corresponding to less than.6 TR/kg carbon. The impact on a vehicle s fuel consumption due to its conventional air conditioning (AC) is a real problem for the automotive industry: about 8% to5% of car or truck fuel consumption is currently attributed to the AC system [8, 9]. At the same time, a vehicle engine could waste around 6% of the input fuel s energy through the water coolant system and exhaust gas. As the price of fuel increases, it becomes obvious that the development of cost effective (capital and running cost) and environmentally friendly mobile AC is needed. Sorption cooling technology is an attractive alternative, as the system could be powered with waste heat from the engine water cooling jacket (generating temperature up to C) and/or from the exhaust gas (generating temperature up to C); the technology could contribute to substantial fuel savings of up to % in a vehicle, therefore reducing gas emissions (CO, CO, NO x, SO x ). In order to make this technology as competitive as conventional mechanical AC technology, it is critical to select an appropriate adsorbent for the generator that leads to high cooling production density and therefore low capital cost. The space heating during the winter and the hot water requirement throughout the year represents a substantial proportion of energy consumption in the EU: in the UK in, it uses 39% []. Therefore improvements in space heating and hot water production efficiency could lead to a significant fuel saving and gas emissions reduction. The gas fired adsorption heat pump is suitable for the replacement of current conventional gas boilers. Once again, the selection of an appropriate carbon-ammonia pair is very important. This paper presents simulation results for various carbon adsorbents using ammonia refrigerant. The results are then used to identify the adsorbent which gives the highest performance, with the prospect of designing better adsorption refrigeration systems for ice making, air conditioning and heat pumping applications. The thermodynamic model used is mainly based on the adsorption equilibrium equations of the adsorbent-refrigerant pair, the heat required (sensible heat and heat of desorption), the cooling production (latent heat of vaporisation of the refrigerant) and heat production (heat of adsorption and the heat of condensation). The simulation results of 5 various activated carbon-ammonia pairs are investigated at typical conditions: T C = 35 C, T E = -5 C and T = 35 C for ice making ; T C = 35 C, T E = C and T = 35 C for air conditioning and T C = 4 C, T E = 5 C and T = 4 C for heat pumping. The driving temperature T3 varies from 8 C up to C. Three cycles are investigated: single bed, two-bed and infinite number of beds. THERMODYNAMIC MODELLING 5

7 A program written in Matlab 7. is used to estimate of the performance (mainly and thermal production) with different cycles (single bed, two-bed with regeneration, and infinite number of beds) at various working conditions. The code is designed to test a large number of pairs providing that the equilibrium equation of the adsorbent-refrigerant pair and the thermophysical properties of both adsorbent and refrigerant are known. Equilibrium equation Experimental porosity tests were carried out for each sample and the data fitted to a modified form of the Dubinin-Radushkevich (D-R) equation: T x = xoexp k Tsat where: x is the ammonia concentration (kg ammonia kg - carbon); T is the carbon temperature (K); x o is the ammonia concentration under saturation conditions (kg ammonia kg - carbon); T sat is the saturation temperature corresponding to the gas pressure (K). Heat input (sensible heat and heat of desorption) The heat input of each bed is calculated by integration of the effective specific heat (dq/dt) along the process path from the initial temperature T to the generating temperature T 3 as described by Meunier []. The specific heat (dq/dt) along the isosteric path (-) is given by: dq dt = C n (T) + pc xc pa where: C pc is the carbon specific heat (J kg - K - ) which varies with temperature and C pa is the specific heat of adsorbed phase (C pa ~49 J kg - K - for ammonia). The specific heat (dq/dt) along the isobaric path (-3) is given by: () () dq dt = C pc (T) + xc pa x + H T (3) p where: H is the heat of desorption given by the following expression: H = R(P,T) A where: R is the gas constant at the system pressure P and temperature T (J kg - K - ); A is the slope of the saturated adsorbate line on the Clapeyron diagram (A=83.4 K) ; T is the carbon temperature (K) ; T sat is the saturation temperature corresponding to the gas pressure (K). Since the ratio T/T sat is constant along an isostere, the temperature at the end of the isosteric pressurisation phase T can be expressed as: (5) T = T T T T sat C T where: T, T C and T E are the bed initial, condensing and evaporating temperatures, respectively (K). E (4) 6

8 Heat rejected (sensible heat and heat of adsorption) A similar technique is also used to estimate the heat rejected by the bed (heat of adsorption and sensible). All previous equations are unchanged except equation (3), that must take into account the cooling effect on the bed of the cold gas coming from the evaporator, and equation (5). Therefore the specific heat (dq/dt) along the isobaric path (4-) is: dq x (6) = C dt pc (T) + xc pa + [ H ( h (T ) h (T ))] gas gas E T In the same manner as equation (5), the temperature at the end of the isosteric depressurisation stage T 4 may be expressed as: (7) T 4 = T T E T 3 where: T 3 is the bed temperature at the end of the desorption phase. c The effective specific heat is calculated for every K temperature interval throughout both the heating and cooling processes. The total heat input to the bed Q input corresponds to the sum of the combined sensible heat and heat of desorption and is calculated by integration from equation () and equation (3) as illustrated in Figure 3. P Figure 3: Effective specific heat vs. temperature for a single bed In the two-bed regenerative cycle as illustrated in Figure 4, the heat rejected Q reg from the adsorbing bed between T 3 and T 6 is transferred across a temperature difference and is recovered in order to preheat the desorbing bed from T to T 5 until it reaches a predetermined temperature difference K T (typical value K). The total heat input is therefore reduced from Q in + Q reg to Q in and the efficiency of the system is improved. In the case of an infinite number of beds and ideal heat transfer the maximum amount of rejected heat Q reject is recovered. The total heat input required will be reduced from Q in to Q in - Q reject as illustrated in Figure 5, All the methods of heat management in sorption refrigeration systems are well documented [, ]. 7

9 Figure 4: Effective specific heat vs. temperature for a two-bed regenerative cycle [] Q reject Q reg Q in Figure 5: Effective specific vs. temperature in an infinite number of beds [] 8

10 Cooling production and c The specific cooling production Q c (kj kg - carbon) is characterized by the latent heat of vaporisation of the ammonia liquid collected during the condensation phase: Q c ( h (T ) h (T )) = x (8) gas where: x is the amount of liquid ammonia collected during the condensation phase (kg ammonia/kg carbon) corresponding to the net concentration change from the higher isostere line (at T ) to the lower isostere line (at T 3 ), and thus the effect of generator void volume is neglected; h gas and h liquid are the specific enthalpies of the gas leaving the evaporator and the condensed liquid (kj kg - ). The cooling production could also be expressed per unit volume of carbon q c (kj m -3 ): q c = ρq (9) c where: ρ is the bulk density of the carbon sample (kg m -3 ) The cooling coefficient of performance c is the ratio of the cooling production to the heat input: () E Q c = Q c input liquid C Heating production and H From the heat balance throughout the full cycle, the relationship between the cooling coefficient of performance c and the heating coefficient of performance H is given by the following expression: = () H c + Since the heating coefficient of performance H is the ratio of the heating production Q H to the heat input, () Q H = Q The heating production Q H is therefore given by: H H input ( c + ) Q input Q = (3) The heating production could also be expressed per unit volume of carbon q H (kj m -3 ): q = ρ (4) H Q H where: ρ is the bulk density of the carbon sample (kg m -3 ) CARBON-AMMONIA PAIR CHARACTERISTICS 9

11 The carbon adsorbents investigated with ammonia refrigerant are mainly coconut shell and coal based types in multiple form: monolithic, granular, compacted granular, fibre, compacted fibre, cloth, compacted cloth and powder. The main properties of each carbon-ammonia pair, namely the Dubinin coefficients (xo, K, n) and density (ρ) that were experimentally measured are summarized in Table. Density Carbon Reference Manufacturer x K n SEE (kg m -3 ) Origin MONOLITHIC CARBON LM Chemviron Carbon Coco Shell LM7 Chemviron Carbon Coco Shell LM8 Chemviron Carbon Coco Shell LM79 Chemviron Carbon Coco Shell SDS ATMI PVDC KOH-AC Chemviron Carbon GRANULAR CARBON 8C Chemviron Carbon Coco Shell 67C Chemviron Carbon Coco Shell C9 Carbomafra SRD35/ Chemviron Carbon Coco Shell SRD35/3 Chemviron Carbon Coco Shell SRD638 Chemviron Carbon Coal SRD639 Chemviron Carbon Coal SRD64 Chemviron Carbon Coal SRD64 Chemviron Carbon Coal CARBON FIBRE & CLOTH ACF CC Chemviron Carbon Polymer ACF CC5 Chemviron Carbon Polymer FM/7 Chemviron Carbon Polymer ACF- Osaka Gas Chemical CARBON POWDER C-3 Westvaco Coal AX- Andersen Coal MSC-3 Kansai Coal COMPACTED SAMPLES 8C Chemviron Carbon Coco Shell SRD35/ Chemviron Carbon Coco Shell FM/7 Chemviron Carbon Polymer ACF- Osaka Gas Chemical Coco Shell = Coconut Shell Table : Carbon adsorbent characteristics SIMULATION RESULTS

12 Ice making applications (T C = 35 C, T E = -5 C and T = 35 C) Monolithic carbon The results are shown in Figure 6. For single bed and two-beds, the sample KOH-AC gives the highest (.46 and.67 for single bed and two-beds, respectively, with a driving temperature of C) while the sample LM gives the lowest (.36 and.55 for single bed and two-beds, respectively, with a driving temperature of C). From single to two-beds, there is an improvement in of up to 3%. Considering the ideal case (an infinite number of beds and ideal heat transfer), the could double (~.75) at lower and medium driving temperature or triple (~.5) at the higher driving temperatures (7 C<T 3 < C) ; the difference in between the monolithic carbons tested is marginal. The cooling production ranges from 5 MJ m -3 to 3 MJ m -3. The KOH-ACammonia pair gives the best cooling production: 66 MJ m -3 (T 3 = C) and 8 MJ m -3 (T 3 = C). The LM-ammonia pair gives the lowest cooling production: 39 MJ m -3 (T 3 = C) and 8 MJ m -3 (T 3 = C). Granular carbon Figure 7 shows the simulation results for various granular carbons. For a single bed and twobeds, the sample SDR 35/ and SRD35/3 give the highest (.46 and.67 for single bed and double bed respectively with a driving temperature of C) while the sample C9 shows the lowest (.34 and.53 for a single bed and two-beds, respectively, with a driving temperature of C). From single to two-beds, there is an improvement in of up to 35%. With the ideal case the could double (~.8) at lower and medium driving temperature or triple (~.3 with SRD35/) or quadruple (~.6 with SRD64) at higher driving temperature (7 C<T 3 < C). The cooling production ranges from 4 MJ m -3 to 6 MJ m -3. The SRD35/-ammonia pair gives the highest cooling production: 36 MJ m -3 (T 3 = C) and 6 MJ m -3 (T 3 = C). The C9-ammonia pair gives the lowest cooling production: 9 MJ m -3 (T 3 = C) and 7 MJ m -3 (T 3 = C). Carbon fibre and cloth The carbon cloth FM/7 and carbon fibre ACF- have fairly similar and best value when considering both single and two-bed cycles:.43 and.63 for single bed and two-beds respectively with a driving temperature of C (Figure 8). From single to two-bed, there is an improvement in of up to 33%. With the ideal case the could double (~.8) at lower and medium driving temperature or triple (~.3 with the carbon cloth FM/7 and carbon fibre ACF-) or quadruple (~.7 with the carbon cloth CC and C5) at higher driving temperature (7 C<T 3 < C). The cooling production varies from MJ m -3 to 9 MJ m -3. The carbon cloth FM/7-ammonia pair gives the highest cooling production: 7 MJ m -3 (T 3 = C) and 83 MJ m -3 (T 3 = C). The carbon fibre ACF-- ammonia pair gives the lowest cooling production: 8 MJ m -3 (T 3 = C) and 5 MJ m -3 (T 3 = C). Carbon powder The carbon powders tested are highly activated as carbon fibre and cloth samples and globally present fairly similar figures of. The best value when considering both single and double bed configuration:.45 and.7 for single bed and double bed, respectively, with a driving temperature of C for the AX- (Figure 9). The cooling production varies from 4 MJ m -3 to MJ m -3. The carbon powder C-3-ammonia pair gives the highest cooling production up to a driving temperature of about 4 C (7 MJ m -3 ) while beyond 4 C the carbon powder AX--ammonia pair is better ( MJ m -3 with T 3 = C). Compacted samples of carbon The of the compacted samples is unchanged since it is mainly driven by ammonia uptake (kg ammonia/kg carbon). However, the cooling production for a given sample is

13 directly proportional to its density. The density improvements are about 6% with SRD3/, 7% with FM/7, 33% with 8C and 65% with ACF-: the cooling production improvement with each compacted sample will follow the same trend. The compacted granular carbon 8C-ammonia pair presents globally the best cooling production with a maximum of 3 MJ m -3 with a driving temperature of C (Figure ). Both and cooling production increase with the bed driving temperature as expected. For the ice making application and regardless of the type of carbon sample, the need to use a high driving temperature beyond C to 4 C depends on the heat source. For free heat source such as solar, agricultural or industrial waste, the cooling production is the main figure of merit and any cooling gained at high driven temperature is appreciated. However, with a gas-fired heat source, the is critical and performance benefits must be significant to justify use of driving temperatures beyond C to 4 C. With the exception of the monolithic SDS (up to 7% increase) no other samples tested seems appropriate for use above 4 C with one bed, as the increases remain marginal (less than 7%). With a two-bed cycle, it will be useful to operate beyond 4 C driving temperature since the increase in could be substantial (up to 5%). Air conditioning applications (T C = 35 C, T E = C and T = 35 C) Monolithic carbon The results are shown in Figure. For a single bed and two-beds, the sample KOH-AC gives the highest (.5 and.85 for single bed and two-bed, respectively, with a driving temperature of C) while the monolithic sample SDS gives the lowest (.39 and.66 for single bed and double bed, respectively, with a driving temperature of C). From single to two-beds, there is an improvement in of up to 4%. Considering the ideal case, the could increase by up to a factor of nine ( Single ~.4 and ideal ~3.6 with LM at T 3 = C). The cooling production ranges from 5 MJ m -3 to 3 MJ m -3. The KOH-AC-ammonia pair gives the highest cooling production: 5 MJ m -3 (T 3 = C) and 36 MJ m -3 (T 3 = C). The LM-ammonia pair gives globally the lowest cooling production: 6 MJ m -3 (T 3 = C) and 4 MJ m -3 (T 3 = C). Granular carbon Figure shows the simulation results with various granular carbons. For a single bed and two-bed, the sample SRD35/ gives the highest (.55 and.87 for single bed and twobeds, respectively, with a driving temperature of C) while the sample SRD64 gives the lowest (.4 and.7 for single bed and two-bed respectively with a driving temperature of C). From single to two-bed, there is an improvement in of up to 37%. In the ideal case the could increase by a factor of ten ( Single ~.4 and ideal ~4. with SRD64 at T 3 = C). The cooling production ranges from 8 MJ m -3 to 6 MJ m -3. The SRD53/-ammonia pair gives the highest cooling production: 9 MJ m -3 (T 3 = C) and 6 MJ m -3 (T 3 = C). The C9-ammonia pair gives the lowest cooling production: 43 MJ m -3 (T 3 = C) and 94 MJ m -3 (T 3 = C). Carbon fibre and cloth The carbon cloth FM/7 and carbon fibre ACF- have similar and highest value are.5 and.84 for single bed and two-bed, respectively, with a driving temperature of C (Figure 3). From single to two-bed, there is an improvement in of up to 4%. In the ideal case the could increase by a factor of ten ( Single ~.43 and ideal ~4.3 with CC at T 3 = C). The cooling production varies from 5 MJ m -3 to 5 MJ m -3. The carbon cloth FM/7-ammonia pair gives the highest cooling production: 6 MJ m -3 (T 3 = C) and MJ m -3 (T 3 = C). The carbon fibre ACF--ammonia pair gives the lowest cooling production: MJ m -3 (T 3 = C) and 38 MJ m -3 (T 3 = C).

14 Carbon powder The highest values of are.45 and.7 for single bed and two-bed, respectively, with a driving temperature of C for the AX- (Figure 4). From single to two-bed, the improvement is up to 36%. With the ideal cycle, the could increase by up to a factor of seven ( Single ~.54 and ideal ~3.6 with AX- at T 3 = C). The cooling production varies from MJ m -3 to 5 MJ m -3. The carbon powder C-3-ammonia pair gives the highest cooling production: 74 MJ m -3 (T 3 = C) and 45 MJ m -3 (T 3 = C). Compacted samples of carbon As with the ice making application, the of the compacted samples is unchanged since it is mainly driven by ammonia uptake (kg ammonia/kg carbon). The cooling production improvement with each compacted sample will also follow the same trend as the density improvement (6% with SRD3/, 7% with FM/7, 33% with 8C and 65% with ACF-). The compacted granular carbon SRD3/-ammonia pair presents the best cooling production with a maximum of 9 MJ m -3 with a driving temperature of C (Figure 5). As with the ice making applications, both and cooling production also increase with the bed driving temperature as expected. Considering a single bed or two-bed cycle and with the exception of the monolithic SDS (up to % of Double bed increase), the variation of is marginal beyond a driving temperature of about 4 C (less than a 7% increase). Unless the heat source is free, it will not be useful to operate beyond a driving temperature of about 4 C unless there is a substantial gain in cooling production (e.g. for the monolithic carbon KOH-AC: about % cooling production gain from 4 C to C driving temperature). Heat pump applications (T C = 4 C, T E = 5 C and T = 4 C) Monolithic carbon For single bed and two-bed cycles, the sample KOH-AC gives the highest (.5 and.7 for single bed and two-bed, respectively, with a driving temperature of C) while the sample SDS gives the lowest (.36 and.56 for single bed and two-bed, respectively, with a driving temperature of C) as shown in Figure 6. From single to two-bed, there is an improvement in of up to 3%. Considering the ideal case, the could double (~3) at higher driving temperature (7 C< T 3 < C); the difference in between the monolithic carbons tested is globally marginal. The heating production ranges from MJ m -3 to 8 MJ m -3. The KOH-AC-ammonia pair gives the highest heating production: 37 MJ m -3 (T 3 = C) and 594 MJ m -3 (T 3 = C) with a two-bed cycle. The LM-ammonia pair gives the lowest heating production: 3 MJ m -3 (T 3 = C) and 38 MJ m -3 (T 3 = C) with a two-bed cycle. Granular carbon Figure 7 shows the simulation results with various granular carbons. For single bed and two-bed, the sample SRD35/ gives the highest (.5 and.73 for single bed and twobed, respectively, with a driving temperature of C) while the sample SRD64 gives the lowest (.38 and.59 for single bed and two-bed, respectively, with a driving temperature of C). From single to two-bed, there is an improvement in of up to 3%. With the ideal case, the could double. The heating production ranges from 4 MJ m -3 to 37 MJ m -3. The SRD35/-ammonia pair gives the highest heating production: 8 MJ m -3 (T 3 = C) and 87 MJ m -3 (T 3 = C) with double bed configuration. The C9- ammonia pair gives the lowest cooling production: 8 MJ m -3 (T 3 = C) and 6 MJ m -3 (T 3 = C) with a two-bed cycle. Carbon fibre and cloth The carbon cloth FM/7 and carbon fibre ACF- have fairly close and the highest values are.46 and.69 for single bed and two-bed, respectively, with a driving temperature 3

15 of C (Figure 8). From single to two-bed, there is an improvement in of up to 4%. With the ideal case, the could double. The heating production varies from 8 MJ m -3 to 3 MJ m -3. The carbon cloth FM/7-ammonia pair gives the highest heating production: 97 MJ m -3 (T 3 = C) and 35 MJ m -3 (T 3 = C) with a two-bed cycle. The carbon fibre ACF--ammonia pair gives the lowest heating production: 3 MJ m -3 (T 3 = C) and 7 MJ m -3 (T 3 = C) with a two-bed cycle. Carbon powder The highest values of are.5 and.8 for single bed and two-bed, respectively, with a driving temperature of C and are obtained with the AX- (Figure 9). The heating production varies from 4 MJ m -3 to 336 MJ m -3. The carbon powder C-3-ammonia pair gives the highest heating production: MJ m -3 (T 3 = C) and 6 MJ m -3 (T 3 = C) with a two-bed cycle. Compacted samples of carbon As with the ice making and air conditioning applications, the of the compacted samples is unchanged since it is mainly driven by ammonia uptake (kg ammonia/kg carbon). The heating production improvement with each compacted sample will also follow approximately the same trend as the density improvement (6% with SRD3/, 7% with FM/7, 33% with 8C and 65% with ACF-). The compacted granular carbon 8C-ammonia pair presents globally the best heating production with maximum values of 5 MJ m -3 and 368 MJ m -3 with a driving temperature of C for single bed and two-bed cycles, respectively (Figure ). Regardless of the type of carbon tested, both and heating production increase with the bed driving temperature as expected since they are all driven by the variation of ammonia uptake in the bed (Figure 6, Figure 7, Figure 8, Figure 9 and Figure ). Globally the varies from to 3.5 while the heating production ranges from 5 MJ m -3 (infinite number of beds) to 8 MJ m -3 (Single bed). Considering a single bed configuration, the variation of is marginal beyond a driving temperature of about C to 4 C (less than a 6% increase). Unless the heat source is free, it will not be useful to operate beyond a driving temperature of about 4 C (with single or two-bed cycles).. With an infinite number of beds (Ideal), the varies little with driving temperature up to about C and has an average value of about.8 which is independent of the sample. The heating production decreases with the number of beds as expected, since increases and less heat is externally rejected during adsorption. With an ideal cycle (infinite number of beds) and for all three applications investigated (ice making, air conditioning and heat pumping), the ideal depends only on the driving temperature (T 3 ) and is independent of the sample until a tipping temperature. At this temperature the starts to increase sharply with increased driving temperature and there is divergence between the s of the carbon samples. The tipping temperature (T tip ) corresponds to the driving temperature at which T = T 4 as illustrated in both Figure and Figure 5 and is given by: T tip TC = T (5) TE This driving temperature defines the driving temperature above which some part of the heat of adsorption can be recovered to provide some of the heat of desorption, therefore reducing the heat input required for the desorption process and causing the rapid increase in the 4

16 above this temperature. This tipping temperature is calculated for the three applications in Table. Ice making Air conditioning Heat pumping T E ( o C) -5 5 T C ( o C) T ( o C) T = T 4 ( o C) T tip ( o C) Table : Tipping temperatures for the three applications Below the tipping temperature, only the sensible heat of the bed can be recovered to provide some of the heat of desorption as illustrated in Figure for carbon 8C in the air conditioning application with a driving temperature of 85 C. The external heat input to the cycle Q input is given by the area of the hatch-marked region which has been split into two sections Q hot and Q hot. The heat input Q hot is required due to the lower ammonia concentration in the bed during the depressurisation stage than during the pressurisation stage of the cycle (and therefore a lower bed heat capacity), and is given by: Q = xc (T T ) (6) hot pa 4 At low driving temperatures, the temperature-concentration characteristics for the carbons is extremely linear as illustrated in Figure for two of the carbon samples. This means that if the gas constant R can be assumed constant at the condensing pressure P C between temperatures T and T 3 then the following simplified expression can be derived for Q hot : (T + T x 3 Qhot R A x + (T3 T Tc ) =. (7) Q input =Q hot +Q hot is therefore directly proportional to the concentration change x. The cooling produced, Q c, is also proportional to the concentration change as given by equation (4) and therefore the is independent of the carbon in this region. Beyond the tipping temperature, the external heat input is illustrated by the hatch-marked area in Figure 3 for the air conditioning application for 8C. The heat of adsorption can be used to provide some of the heat of desorption between temperatures T and T 4. It can be seen from the figure that the heat of adsorption is much lower in proportion to the heat of desorption for MSC-3 in this region than for 8C. Therefore, 8C has a higher ideal above the tipping temperature. This is due to the lower rate of change of concentration with temperature for MSC-3 during desorption compared to adsorption in the region between T and T 4, as illustrated in Figure 4. CONCLUSION The simulation results of various activated carbon ammonia pairs are presented at typical conditions: T C = 35 C, T E = -5 C and T = 35 C for ice making; T C = 35 C, T E = C and T ) c pa 5

17 = 35 C for air conditioning and T C = 4 C, T E = 5 C and T = 4 C for heat pumping. The driving temperature T 3 varies from 8 C up to C (8 C < T 3 < C). The simulations were carried out for three configurations of carbon-ammonia bed: single bed, double beds and infinite number of beds. The carbon absorbents investigated are mainly coconut shell and coal based types in various shapes and sizes (monolithic, granular, compacted granular, fibre, compacted fibre, cloth, compacted cloth and powder). Considering a double bed configuration at a low driving temperature of C (solar or low grade waste heat source driven machines) and a high driving temperature of C (engine exhaust gas or gas-fired driven machines), the best performances are summarized in Table 3 and Table 4 respectively: Samples q c (MJ m -3 ) Ice making Air conditioning Heat pumping q c q H (MJ m -3 ) (MJ m -3 ) KOH-AC (monolithic) SRD35/ (Granular) FM/7 (Fibre) C-3 (Powder) SRD35/ (Compacted) Table 3: Best performance with two-bed configuration at driving temperature of C. 6

18 Samples q c (MJ m -3 ) Ice making Air conditioning Heat pumping q c q H (MJ m -3 ) (MJ m -3 ) KOH-AC (monolithic) SRD35/ (Granular) FM/7 (Fibre) (Powder) (Compacted) AX- 8C 3 AX-.7 8C.58 C-3 45 SRD35/ 9 C-3.88 SRD35/.87 AX- 65 8C 368 AX-.8 8C.6 Table 4: Best performance with two-bed configuration at driving temperature of C. ACKNOWLEDGEMENTS This research is supported by the combined following grants: ) EU SOCOOL project under the grant ENK5-CT-63. ) EU TOPMACS project under the grant TST4-CT ) EPSRC (UK-Engineering & Physical Sciences Research Council) under the grant EP/C388/. The samples carbon were donated by the following individuals and companies: 4) Chemviron Carbons Ltd (Lockett Road, Lancashire WN4 8DE, UK). 5) ATMI Inc (7 Commerce Drive Danbury, CT 68, USA). 6) WESTVACO (Charleston, SC 943, USA). 7) Pr B.B. Saha (Institute for Materials Chemistry and Engineering, Kyushu University, Kasuga-koen 6-, Kasuga-shi, Fukuoka , Japan). 7

19 REFERENCES [] R. E. Critoph - Towards one tonne per day solar ice maker. Proc. IVth World Renewable Energy Congress, Denver (US), 996, pp [] R. E. Critoph, Z. Tamainot-Telto and E. Munyebvu - Solar Sorption Refrigerator - Renewable Energy, (4) (997), pp [3] Z. F. Li and K. Sumathy - A solar powered ice maker with the solid adsorption pair of activated carbon and methanol - Int J Energy Res 3 (999), pp [4] A. Boubakri, J. J. Guilleminot and F. Meunier - Adsorptive solar powered ice maker: experiments and model - Solar Energy 69 (3) (), pp [5] A. P. F. Leite and M. Daguenet - Performance of a new solid adsorption ice maker with solar energy regeneration. Energy Conversion Mgmt 4 (), pp [6] M. Li, R. Z. Wang, H. L. Luo, L. L. Wnag and H. B. Huang, Experiments of a solar at plate hybrid system with heating and cooling. Applied Thermal Engineering (), pp [7] F. Buchter, P. H. Dind and M. Pons - An experimental solar-powered adsorptive refrigeration tested in Burkina Faso. Int. J Refrigeration 6 (3), pp [8] S.J. Metcalf, Z. Tamainot-Telto and R. E. Critoph - Compact sorption generator for car and truck using waste-heat, rst European Mobile Air Conditioning Workshop, Paper No 5A99, Torino, Italy, November 5. [9] R. E. Critoph, S. J. Metcalf, and Z. Tamainot-Telto - Compact plate adsorbers for car air conditioning applications, International Heat Powered Cycles Conference (HPC 6), Paper No 64, Newcastle, UK, September 6. [] S. J. Metcalf Gas-fired adsorption heat pump for domestic gas boiler replacement - International Heat Powered Cycles Conference (HPC 6), Paper No 637, Newcastle, UK, September 6. [] F. Meunier Second law analysis of a solid adsorption heat pump operating on reversible cascade cycles: application to the zeolite-water pair, Heat Recovery Systems, 5 (985), pp [] R. E. Critoph Adsorption refrigerators and heat pumps Carbon materials for advanced technologies, Edited by Timothy D. Burchell, Chap., pp ,999, ISBN

20 9

21 .5 Ice Making: Single bed (Monolithic carbon) Ice Making: Two-bed (Monolithic carbon) LM LM7 LM8 LM79 SDS KOH-AC.6.4 LM LM7 LM8. LM79 SDS KOH-AC.5.5 Ice Making: Ideal (Monolithic carbon) LM LM7 LM8 LM79 SDS KOH-AC Cooling density (MJ m -3 ) Ice Making: Cooling (Monolithic carbon) LM LM7 LM8 LM79 SDS KOH-AC Figure 6: Ice making using monolithic carbon

22 Ice Making: Single bed (Granular carbon) 8C 67C C9 SRD35/ SRD35/3 SRD638 SRD639 SRD64 SRD Ice Making: Two-Bed (Granular carbon) 8C.4 67C C9 SRD35/. SRD35/3 SRD638 SRD SRD64 8 Driving Temperature ( o SRD64 C).5 Ice Making: Ideal (Granular carbon) 8C 67C C9 SRD35/.5 SRD35/3 SRD638 SRD639 SRD64 SRD64 Cooling density (MJ m -3 ) 5 5 Ice Making: Cooling (Granular carbon) 8C 67C C9 SRD35/ SRD35/3 SRD638 SRD639 SRD64 SRD64 Figure 7: Ice making using granular carbon

23 .5 Ice Making: Single bed (Fibre & Cloth carbon) Ice Making: Two-bed (Fibre & Cloth carbon) CC CC5 FM/7 ACF-.4. CC CC5 FM/7 ACF-.5.5 Ice Making: Ideal (Fibre & Cloth carbon) CC CC5 FM/7 ACF- Cooling density (MJ m -3 ) 5 5 Ice Making: Cooling (Fibre & Cloth carbon) CC CC5 FM/7 ACF- Figure 8: Ice making using carbon fibre and cloth

24 .5 Ice Making: Single bed (Powder carbon) Ice Making: Two-bed (Powder carbon) C-3 AX- MSC-3. C-3 AX- MSC-3.5 Ice Making: Ideal (Powder carbon) 5 Ice Making: Cooling (Powder carbon).5 C-3 AX- MSC-3 Cooling density (MJ m -3 ) 5 C-3 AX- MSC-3 Figure 9: Ice making using carbon powder 3

25 .5 Ice Making: Single bed (Compacted carbon) Ice Making: Two-bed (Compacted carbon) C. SRD35/ FM/7 ACF-. 8C SRD35/ FM/7 ACF-.5 Ice Making: Ideal (Compacted carbon) 5 Ice Making: Cooling (Compacted carbon).5 8C SRD35/ FM/7 ACF- Cooling density (MJ m -3 ) 5 8C SRD35/ FM/7 ACF- Figure : Ice making using compacted carbon 4

26 .8.6 Air Conditioning: Single bed (Monolithic carbon) LM LM7 LM8 LM79 SDS KOH-AC.8.6 Air Conditioning: Two-bed (Monolithic carbon) LM LM7 LM8 LM79 SDS KOH-AC 4 Air Conditioning: Ideal (Monolithic carbon) 3 Air Conditioning: Cooling (Monolithic carbon) 3 LM LM7 LM8 LM79 SDS KOH-AC Cooling density (MJ m -3 ) 5 5 LM LM7 5 LM8 LM79 SDS KOH-AC 8 Figure : Air conditioning using monolithic carbon 5

27 Air Conditioning: Single bed (Granular carbon) Air Conditioning: Two-Bed (Granular carbon) C 67C. C9 SRD35/ SRD35/3 SRD SRD639 8 SRD64 SRD64 4 Air Conditioning: Ideal (Granular carbon).6 8C 67C.4 C9 SRD35/ SRD35/3. SRD638 SRD639 SRD SRD Air Conditioning: Cooling (Granular carbon) 3 8C 67C C9 SRD35/ SRD35/3 SRD638 SRD SRD64 8 SRD64 Cooling density (MJ m -3 ) 8C 67C C9 5 SRD35/ SRD35/3 SRD638 SRD639 SRD SRD64 8 Figure : Air conditioning using granular carbon 6

28 .8 Air Conditioning: Single bed (Fibre & Cloth carbon) CC CC5 FM/7 ACF-.8 Air Conditioning: Two-bed (Fibre & cloth carbon) CC CC5 FM/7 ACF- 4 3 Air Conditioning: Ideal (Fibre & cloth carbon) CC CC5 FM/7 ACF- Cooling density (MJ m -3 ) 5 5 CC CC5 FM/7 ACF- Figure 3: Air conditioning using carbon fibre and cloth 7

29 .8 Air Conditioning: Single bed (Powder carbon) C-3 AX- MSC-3.8 Air Conditioning: Two-bed (Powder carbon) C-3 AX- MSC Air Conditioning: Ideal (Power carbon) C-3 AX- MSC-3 Cooling density (MJ m -3 ) 5 5 C-3 AX- MSC-3 Figure 4: Air conditioning using carbon powder 8

30 Air Conditioning: Cooling (Compacted carbon).8 Air Conditioning: Single bed (Compacted carbon) 8C SRD35/ FM/7 ACF-.8 Air Conditioning: Two-bed (Compacted carbon) C SRD35/ FM/7 ACF- 4 3 Air Conditioning: Ideal (Compacted carbon) 8C SRD35/ FM/7 ACF- Cooling density (MJ m -3 ) 5 5 8C SRD35/ FM/7 ACF- Figure 5: Air conditioning using carbon compacted 9

31 Heat Pump: Single bed (Monolithic carbon) LM LM7.8 LM8 LM79 SDS.6 KOH-AC.4 Heat Pump: Two-bed (Monolithic carbon) LM LM7.8 LM8 LM79 SDS.6 KOH-AC.4 Heat Pump: Ideal (Monolithic carbon) 4 LM 3.5 LM7 LM8 LM79 3 SDS KOH-AC Heating density (MJ m -3 ) Heat Pump: Single bed (Monolithic carbon) 8 LM LM7 LM8 6 LM79 SDS KOH-AC 4 Heating density (MJ m -3 ) Heat Pump: Two-bed (Monolithic carbon) 8 LM LM7 LM8 6 LM79 SDS KOH-AC 4 Heating density (MJ m -3 ) Heat Pump: Ideal (Monolithic carbon) 8 LM LM7 LM8 6 LM79 SDS KOH-AC 4 Figure 6: Heat pump using monolithic carbon 3

32 Heating density (MJ m -3 ) Heat Pump: Single bed (Granular carbon) C 67C C9 SRD35/ SRD35/3 SRD638 SRD639 SRD SRD64 8 8C 67C C9 SRD35/ SRD35/ SRD638 Driving Temperature ( o SRD639 C) SRD64 SRD64 Heat Pump: Single bed (Granular carbon) 4 Heating density (MJ m -3 ) Heat Pump: Two-Bed (Granular carbon).8.6 8C.4 67C C9 SRD35/. SRD35/3 SRD638 SRD SRD64 8 Driving Temperature ( o SRD64 C) 8C Heat Pump: 67CTwo-Bed (Granular carbon) 4 C9 SRD35/ SRD35/3 3 SRD638 SRD639 SRD64 SRD64 Heating density (MJ m -3 ) Heat Pump: Ideal(Granular carbon) 8C 67C C9 SRD35/ SRD35/3 SRD638 SRD639 SRD64 SRD64 Heat Pump: Ideal(Granular carbon) 8C 67C C9 SRD35/ SRD35/3 SRD638 SRD639 SRD64 SRD64 Figure 7: Heat pump using granular carbon 3

33 Heat Pump: Single bed (Fibre & cloth carbon) CC CC5.8 FM/7 ACF Heat Pump: Two-bed (Fibre & cloth carbon) CC CC5.8 FM/7 ACF Heat Pump: Ideal (Fibre & cloth carbon) 4 CC 3.5 CC5 FM/7 ACF Heating density (MJ m -3 ) Heat Pump: Single bed (Fibre & cloth carbon) 4 CC CC5 FM/7 3 ACF- Heating density (MJ m -3 ) Heat Pump: Two-bed (Fibre & cloth carbon) 4 CC CC5 FM/7 3 ACF- Heating density (MJ m -3 ) Heat Pump: Ideal (Fibre & cloth carbon) 4 CC CC5 FM/7 3 ACF- Figure 8: Heat pump using carbon fibre and cloth 3

34 Heat Pump: Single bed (Powder carbon) C-3 AX-.8 MSC-3 Heat Pump: Two-bed (Powder carbon) C-3 AX-.8 MSC Heat Pump: Ideal (Powder carbon) C-3 AX- MSC Heating density (MJ m -3 ) Heat Pump: Single bed (Powder carbon) 4 C-3 AX- MSC-3 3 Heating density (MJ m -3 ) Heat Pump: Two-bed (Powder carbon) 4 C-3 AX- MSC-3 3 Heating density (MJ m -3 ) 4 3 Heat Pump: Ideal (Powder carbon) C-3 AX- MSC-3 Figure 9: Heat pump using carbon powder 33

35 Heat Pump: Single bed (Compacted carbon) 8C SRD35/.8 FM/7 ACF-.6 Heat Pump: Two-bed (Compacted carbon) 8C SRD35/.8 FM/7 ACF-.6 Heat Pump: Ideal (Compacted carbon) 4 8C 3.5 SRD35/ FM/7 ACF Heating density (MJ m -3 ) Heat Pump: Single bed (Compacted carbon) 4 3 8C SRD35/ FM/7 ACF- Heating density (MJ m -3 ) Heat Pump: Two-bed (Compacted carbon) 4 3 8C SRD35/ FM/7 ACF- Heating density (MJ m -3 ) Heat Pump: Ideal (Compacted carbon) 4 8C SRD35/ 3 FM/7 ACF- Figure : Heat pump using compacted carbon 34

36 Figure : Bed specific heat capacity for 8C in the air conditioning application with a driving temperature of 85 C 35

37 Concentration (kg kg - ) Temperature ( o C) 8C MSC-3 Figure : Temperature-concentration for two carbons during desorption in the air conditioning application at low driving temperature 36

38 Figure 3: Bed heat capacity for adsorption and desorption for 8C and MSC-3 in the air conditioning application with a driving temperature of 95 C. 37

39 Figure 4: Concentration profiles during adsorption and desorption for 8C and MSC-3 for the air conditioning application with a driving temperature of C. 38